In 1974, Folkman proposed that for tumors to grow beyond a critical size and to spread to form metastases, they must recruit endothelial cells from the surrounding stroma to form their own endogenous microcirculation in a process termed angiogenesis (Folkman J. (1974) Adv Cancer Res. 19; 331). The new blood vessels induced by tumor cells as their life-line of oxygen and nutrients also provide exits for cancer cells to spread to other parts of the body. Inhibition of this process has been shown to effectively stop the proliferation and metastasis of solid tumors. A drug that specifically inhibits this process is known as an angiogenesis inhibitor.
Having emerged as a promising new strategy for the treatment of cancer, the anti-angiogenesis therapy (“indirect attack”) has several advantages over the “direct attack” strategies. All the “direct attack” approaches such as using DNA damaging drugs, antimetabolites, attacking the RAS pathway, restoring p53, activating death programs, using aggressive T-cells, injecting monoclonal antibodies and inhibiting telomerase, etc., inevitably result in the selection of resistant tumor cells. Targeting the endothelial compartment of tumors as in the “indirect attack”, however, should avoid the resistance problem because endothelial cells do not exhibit the same degree of genomic instability as tumor cells. Moreover, anti-angiogenic therapy generally has low toxicity due to the fact that normal endothelial cells are relatively quiescent in the body and exhibit an extremely long turnover. Finally since the “indirect attack” and “direct attack” target different cell types, there is a great potential for a more effective combination therapy.
More than 300 angiogenesis inhibitors have been discovered, of which about 31 agents are currently being tested in human trials in treatment of cancers (Thompson, et al., (1999) J Pathol 187, 503). TNP-470, a semisynthetic derivative of fumagillin of Aspergillus fuigatus, is among the most potent inhibitors of angiogenesis. It acts by directly inhibiting endothelial cell growth and migration in vitro and in vivo (Ingber et al. (1990) Nature 348, 555). Fumagillin and TNP-470, have been shown to inhibit type 2 methionine aminopeptidase (hereinafter MetAP2) by irreversibly modifying its active site. The biochemical activity of fumagillin analogs has been shown to correlate to their inhibitory effect on the proliferation of human umbillical vein endothelial cells (HUVEC). Although the mechanism of the selective action of fumagillin and related compounds on MetAP2-mediated endothelial cell cytostatic effect has not yet been established, possible roles of MetAP2 in cell proliferation have been suggested.
First, hMetAP-2-catalyzed cleavage of the initiator methionine of proteins could be essential for releasing many proteins that, after myristoylation, function as important signaling cellular factors involved in cell proliferation. Proteins known to be myristoylated include the src family tyrosine kinases, the small GTPase ARF, the HIV protein nef and the α subunit of heterotrimeric G proteins. A recently published study has shown that the myristoylation of nitric oxide synthase, a membrane protein involved in cell apoptosis, was blocked by fumagillin (Yoshida, et al. (1998) Cancer Res. 58(16), 3751). This is proposed to be an indirect outcome of inhibition of MetAP2-catalyzed release of the glycine-terminal myristoylation substrate. Alternatively, MetAP enzymes are known to be important to the stability of proteins in vivo according to the “N-end rule” which suggests increased stability of methionine-cleaved proteins relative to their N-terminal methionine precursors (Varshavsky, A (1996) Proc. Natl. Acad. Sci. U.S.A. 93, 12142). Inhibition of hMetAP2 could result in abnormal presence or absence of some cellular proteins critical to the cell cycle.
Methionine aminopeptidases (MetAP) are ubiquitously distributed in all living organisms. They catalyze the removal of the initiator methionine from newly translated polypeptides using divalent metal ions as cofactors. Two distantly related MetAP enzymes, type 1 and type 2, are found in eukaryotes, which at least in yeast, are both required for normal growth; whereas only one single MetAP is found in eubacteria (type 1) and archaebacteria (type 2). The N-terminal extension region distinguishes the methionine aminopeptidases in eukaryotes from those in procaryotes. A 64-amino acid sequence insertion (from residues 381 to 444 in hMetAP2) in the catalytic C-terminal domain distinguishes the MetAP-2 family from the MetAP-1 family. Despite the difference in the gene structure, all MetAP enzymes appear to share a highly conserved catalytic scaffold termed “pita-bread” fold (Bazan, et al. (1994) Proc. Natl. Acad. Sci. U.S.A. 91, 2473), which contains six strictly conserved residues implicated in the coordination of the metal cofactors.
Mammalian type 2 methionine aminopeptidase has been identified as a bifunctional protein implicated by its ability to catalyze the cleavage of N-terminal methionine from nascent polypeptides (Bradshaw, et al (1998) Trends Biochem. Sci. 23, 263) and to associate with eukaryotic initiation factor 2α (eIF-2α) to prevent its phosphorylation (Ray, et al. (1992) Proc. Natl. Acad. Sci. U.S.A. 89, 539). Both the genes of human and rat MetAP2 were cloned and have shown 92% sequence identity (Wu,. et al. (1993) J. Biol. Chem. 268, 10796; Li, X. & Chang, Y.-H. (1996) Biochem. & Biophys. Res. Comm. 227, 152). The N-terminal extension in these enzymes is highly charged and consists of two basic polylysine blocks and one aspartic acid block, which has been speculated to be involved in the binding of eIF-2α (Gupta, et al. (1993) in Translational Regulation of Gene Expression 2 (Ilan, J., Ed.), pp. 405-431, Plenum Press, New York).
The anti-angiogenic compounds, fumagillin and its analogs, have been shown to specifically block the exo-aminopeptidase activity of hMetAP2 without interfering with the formation of the hMetAP2: eIF2α complex (Griffith, et al., (1997) Chem. Biol. 4, 461; Sin, et al. (1997) Proc. Natl. Acad. Sci. U.S.A. 94, 6099). Fumagillin and its analogs inactivate the enzymatic activity of hMetAP2 with a high specificity, which is underscored by the lack of effect of these compounds on the closely related type 1 methionine aminopeptidase (MetAP1) both in vitro and in vivo in yeast (Griffith, et al., (1997) Chem. Biol. 4, 461; Sin, et al. (1997) Proc. Natl. Acad Sci. USA. 94, 6099). The extremely high potency (IC50<1 nM) of these inhibitors appears to be due to the irreversible modification of the active site residue, His231, of hMetAP2 (Liu, et al. (1998) Science 282, 1324). Disturbance of MetAP2 activity in vivo impairs the normal growth of yeast (Griffith, et al., (1997) Chem. Biol. 4, 461; Sin, et al. (1997) Proc. Natl. Acad. Sci. U.S.A. 94, 6099; In-house data) as well as Drosophila (Cutforth & Gaul (1999) Mech. Dev. 82, 23). Most significantly, there appears to be a clear correlation between the inhibition effect of flimagillin related compounds against the enzymatic activity of hMetAP2 in vitro and the suppression effect of these compounds against tumor-induced angiogenesis in vivo (Griffith, et al., (1997) Chem. Biol. 4, 461).
Cancer is the second leading cause of death in the U.S., exceeded only by heart disease. Despite recent successes in therapy against some forms of neoplastic disease, other forms continue to be refractory to treatment. Thus, cancer remains a leading cause of death and morbidity in the United States and elsewhere (Bailar and Gornik (1997) N Engl J Med 336, 1569). Inhibition of hMetAP2 provides a promising mechanism for the development of novel anti-angiogenic agents in the treatment of cancers. It has now been discovered that compounds of formulae (I) and (IA) are effective inhibitors of hMetAP2, and thus would be useful in treating conditions mediated by hMetAP2.